Contact UsGet started

Updating an instance group under load

Written by
Updated at November 14, 2024

Using this step-by-step guide, you will configure an instance group and check its operation when updating the configuration. To do this:

  1. Prepare your cloud.
  2. Prepare the environment.
  3. Create an instance group from a Container Optimized Image.
  4. Create a load on an instance.
  5. Update the instance group under load.
  6. Stop the load and get the results.

If you no longer need the resources you created, delete them.

Prepare your cloud

  1. If you don't have the Yandex Cloud command line interface yet, install it.

  2. For the scripts from the tutorial to run correctly, download and install the jq utility.

  3. To check auto scaling, install the wrk benchmarking tool.

The cost of support for the Yandex Cloud instance group includes a fee for:

Prepare the environment

  1. Create a service account named for-load and assign it the editor role:

    1. In the management console, select the folder where you want to create your service account.
    2. In the list of services, select Identity and Access Management.
    3. Click Create service account.
    4. In the Name field, specify for-load.
    5. Click Add role and select the editor role.
    6. Click Create.

    1. Create a service account:

      The folder specified in the CLI profile is used by default. You can specify a different folder using the --folder-name or --folder-id parameter.

      yc iam service-account create --name for-load

      Command result:

      id: ajeab0cnib1p********
folder_id: b0g12ga82bcv********
created_at: "2021-02-09T17:31:32.561702Z"
name: for-load

    2. Assign the role to the service account:

      yc resource-manager folder add-access-binding b0g12ga82bcv******** \
  --role editor \
  --subject serviceAccount:ajeab0cnib1p********
    1. Create a service account named for-load:
      Use the create REST API method for the ServiceAccount resource or the ServiceAccountService/Create gRPC API call.
    2. Assign the service account the editor role for the current folder:
      Use the setAccessBindings REST API method for the Folder resource or the FolderService/SetAccessBindings gRPC API call.

  2. Create a network named yc-auto-network and subnets in two availability zones:

    1. In the management console, select the folder where you want to create a network.
    2. In the list of services, select Virtual Private Cloud.
    3. Click Create network.
    4. In the Name field, specify yc-auto-network.
    5. Select Create subnets.
    6. Click Create network.

    1. Create a network:

      yc vpc network create --name yc-auto-network

      Command result:

      id: enpabce123hd********
folder_id: b0g12ga82bcv********
created_at: "2021-02-09T17:33:32.561702Z"
name: yc-auto-network

    2. Create a subnet in the ru-central1-a availability zone:

      yc vpc subnet create --network-id enpabce123hd******** --range 192.168.1.0/24 --zone ru-central1-a

      Command result:

      id: e1lnabc23r1c********
folder_id: b0g12ga82bcv********
created_at: "2021-02-09T17:34:32.561702Z"
network_id: enpabce123hd********
zone_id: ru-central1-a
v4_cidr_blocks:
- 192.168.1.0/24

    3. Create a subnet in the ru-central1-b availability zone:

      yc vpc subnet create --network-id enpabce123hd******** --range 192.168.2.0/24 --zone ru-central1-b

      Command result:

      id: b1csa2b3clid********
folder_id: b0g12ga82bcv********
created_at: "2021-02-09T17:35:32.561702Z"
network_id: enpabce123hd********
zone_id: ru-central1-b
v4_cidr_blocks:
- 192.168.2.0/24
    1. Create a network:
      Use the create REST API method for the Network resource or the NetworkService/Create gRPC API call.
    2. Create subnets in the ru-central1-a and ru-central1-b availability zones:
      Use the create REST API method for the Subnet resource or the SubnetService/Create gRPC API call.

Create an instance group from a Container Optimized Image

  1. All instance groups are created from Container Optimized Image. Each instance contains a Docker container running a web server that emulates the service utilization.

    1. In the management console, select a folder to host your instance group.
    2. In the list of services, select Compute Cloud.
    3. In the left-hand panel, select Instance groups.
    4. Click Create group of virtual machines.
    5. Under Basic parameters:
      • In the Name field, specify group-for-load.
      • Select Service account for-load.
    6. Under Allocation, select ru-central1-a and ru-central1-b in the Availability zone field.
    7. Under Instance template, click Define and do the following in the window that opens:
      • Under Boot disk image, select the Container Solution tab.
      • Click Configure.
      • In the Docker container settings window:
        • In the Name field, specify nginx.
        • In the Docker image field, click Enter link and enter cr.yandex/yc/demo/autoscaling-example-app:v1.
        • Click Apply.
      • Under Disks:
        • For the boot disk, specify the Size of 30 GB.
      • Under Network settings:
        • In the Network field, select for-load.
      • Under Access:
        • In the Service account field, select for-load.
        • In the Login field, enter the name of the user to be created on the VM.
        • In the SSH key field, paste the contents of the public key file.
      • Click Save.
    8. Under Allow when creating and updating:
      • In the Reduce below target value field, enter 4.
    9. Under Scaling:
      • In the Type field, select Fixed.
      • Set Size to 6.
    10. Under Integration with Network Load Balancer:
      • Enable Create target group.
      • In the Name of the target group field, specify load-generator.
    11. Click Create.

    1. Find out the ID of the latest version of the public Container Optimized Image.

      A Container Optimized Image in a Yandex Container Registry registry may be updated and changed with new releases. This will not automatically update the VM image to the latest version. To create an instance group with the latest Container Optimized Image version, you should check its availability yourself:

      yc compute image get-latest-from-family container-optimized-image --folder-id standard-images

      Command result:

      id: fd8iv792kira********
folder_id: standard-images
created_at: "2021-01-29T13:30:22Z"
...
status: READY
os:
  type: LINUX

    2. Save the specification of the instance group with network load balancer to the specification.yaml file:

      name: group-for-load # The instance group name that must be unique within the folder.
service_account_id: ajeab0cnib1p******** # Service account ID.
allocation_policy: # VM allocation policy in the group.
   zones:
     - zone_id: ru-central1-a
     - zone_id: ru-central1-b
instance_template:
  service_account_id: ajeab0cnib1p******** # ID of the service account for access to private Docker images.
  platform_id: standard-v1 # Platform ID.
  resources_spec:
    memory: 2G # Amount of memory (RAM).
    cores: 2 # Number of processor cores (vCPUs).
  boot_disk_spec:
    mode: READ_WRITE # Disk access mode (read and write).
    disk_spec:
      image_id: <ID_of_the_latest_COI_version>
    type_id: network-ssd # Disk type.
    size: 30G # Disk size.
  network_interface_specs:
   - network_id: enplhg4nncc7******** # Network ID.
     subnet_ids:
       - e1lnabc23r1c******** # Subnet ID.
       - b1csa2b3clid********
     primary_v4_address_spec: {
       one_to_one_nat_spec: {
         ip_version: IPV4 # IPv4 specification for public access to the VM.
       }
     }
  metadata: # Values to be delivered to the VM metadata.
    docker-container-declaration: |- # Key in the VM metadata that is used with the Docker Container specification.
      spec:
        containers:
          - image: cr.yandex/yc/demo/autoscaling-example-app:v1
            name: nginx
            securityContext:
              privileged: false
            restartPolicy: Always
            tty: false
            stdin: false
    ssh-keys: | # Parameter for sending an SSH key to the VM.
      yc-user:ssh-ed25519 ABC...d01 user@desktop.ru # Username for the VM connection.
deploy_policy: # Deployment policy for instances in the group.
  max_unavailable: 4
scale_policy: # Scaling policy for instances in the group.
  fixed_scale:
    size: 6
load_balancer_spec: # Network load balancer.
  target_group_spec:
    name: load-generator

      Note

      You can provide an SSH key in the VM metadata using the ssh-keys parameter or in the user-data string with the user metadata. This tutorial uses the first option.

    3. Create an instance group named group-for-load using the specification.yaml specification:

      yc compute instance-group create --file=specification.yaml

      Command result:

      done (2m18s)
id: cl0kabcquk1g********
folder_id: b0g12ga82bcv********
...
service_account_id: ajeab0cnib1p********
status: ACTIVE
application_load_balancer_state: {}

    1. Get the ID of the latest container-optimized-image version in the standard-images family:
      Use the getLatestByFamily REST API method for the Image resource or the ImageService/GetLatestByFamily gRPC API call.

    2. Create an instance group based on the following specification:

      name: group-for-load # The instance group name that must be unique within the folder.
service_account_id: ajeab0cnib1p******** # Service account ID.
allocation_policy: # VM allocation policy in the group.
   zones:
     - zone_id: ru-central1-a
     - zone_id: ru-central1-b
instance_template:
  service_account_id: ajeab0cnib1p******** # ID of the service account for access to private Docker images.
  platform_id: standard-v1 # Platform ID.
  resources_spec:
    memory: 2G # Amount of memory (RAM).
    cores: 2 # Number of processor cores (vCPUs).
  boot_disk_spec:
    mode: READ_WRITE # Disk access mode (read and write).
    disk_spec:
      image_id: <ID_of_the_latest_COI_version>
    type_id: network-ssd # Disk type.
    size: 30G # Disk size.
  network_interface_specs:
   - network_id: enplhg4nncc7******** # Network ID.
     subnet_ids:
       - e1lnabc23r1c******** # Subnet ID.
       - b1csa2b3clid********
     primary_v4_address_spec: {
       one_to_one_nat_spec: {
         ip_version: IPV4 # IPv4 specification for public access to the VM.
       }
     }
  metadata: # Values to be delivered to the VM metadata.
    docker-container-declaration: |- # Key in the VM metadata that is used with the Docker Container specification.
      spec:
        containers:
          - image: cr.yandex/yc/demo/autoscaling-example-app:v1
            name: nginx
            securityContext:
              privileged: false
            restartPolicy: Always
            tty: false
            stdin: false
    ssh-keys: | # Parameter for sending an SSH key to the VM.
      yc-user:ssh-ed25519 ABC...d01 user@desktop.ru # Username for the VM connection.
deploy_policy: # Deployment policy for instances in the group.
  max_unavailable: 4
scale_policy: # Scaling policy for instances in the group.
  fixed_scale:
    size: 6
load_balancer_spec: # Network load balancer.
  target_group_spec:
    name: load-generator

      Use the createFromYaml REST API method for the InstanceGroup resource or the InstanceGroupService/CreateFromYaml gRPC API call.

  2. Make sure that the instance group was created:

    1. In the management console, select the folder where you created the instance group.
    2. In the list of services, select Compute Cloud.
    3. In the left-hand panel, select Instance groups.
    4. Click the group-for-load instance group name.
    yc compute instance-group list-instances group-for-load

    Command result:

    +----------------------+---------------------------+-----------------+-------------+----------------------+----------------+
|     INSTANCE ID      |           NAME            |   EXTERNAL IP   | INTERNAL IP |        STATUS        | STATUS MESSAGE |
+----------------------+---------------------------+-----------------+-------------+----------------------+----------------+
...
| fhmab0cfsfd1******** | cl0kabcquk1g********-oxig | 178.154.226.108 | 10.130.0.8  | RUNNING_ACTUAL [49m] |                |
| epdabchpdef0******** | cl0kabcquk1g********-aqyg | 130.193.40.55   | 10.129.0.20 | RUNNING_ACTUAL [43m] |                |
+----------------------+---------------------------+-----------------+-------------+----------------------+----------------+

    To view the list of created instance groups, use the list REST API method for the InstanceGroup resource or the InstanceGroupService/List gRPC API call.

Add a network load balancer to the created instance group

  1. Create a network load balancer named load-generator and connect it to the instance group you created:

    1. In the management console, select the folder you want to create a load balancer in.
    2. In the list of services, select Network Load Balancer.
    3. Click Create a network load balancer.
    4. In the Name field, specify load-generator.
    5. In the Public address field, select Auto.
    6. Under Listeners, click Add listener.
    7. In the window that opens:
      • In the Name field, specify http.
      • In the Port field, enter 80: the balancer will use this port to accept incoming traffic.
      • In the Target port field, enter 80: the balancer will redirect traffic to this port.
      • Click Add.
    8. Under Target groups, click Add target group.
    9. In the Target group list, select load-generator.
    10. Under Health check, click Configure:
      • In the window that opens, enter the Path /hello: the load balancer will use this path to send health check requests to target group VMs.
      • Click Apply.
    11. Click Create.

    1. Get the ID of the load-generator target group:

      yc load-balancer target-group get load-generator | grep "^id"

      Command result:

      id: enpsa475ej51********

    2. Create a network load balancer:

      yc load-balancer network-load-balancer create \
  --name load-generator \
  --listener external-ip-version=ipv4,name=http,port=80,protocol=tcp,target-port=80 \
  --target-group healthcheck-http-port=80,healthcheck-http-path=/hello,target-group-id=<target_group_ID>

      Command result:

      done (14s)
id: b0ruab1ccvpd********
folder_id: b1csa2b3clid********
...
    healthy_threshold: "2"
    http_options:
      port: "80"
      path: /hello
    1. Create a load balancer using the create REST API method for the NetworkLoadBalancer resource or the NetworkLoadBalancerService/Create gRPC API call.
    2. Add a listener to the load balancer using the addListener REST API method for the NetworkLoadBalancer resource or the NetworkLoadBalancerService/AddListener gRPC API call.
    3. Connect the target group to the load balancer using the attachTargetGroup REST API method for the NetworkLoadBalancer resource or the NetworkLoadBalancerService/AttachTargetGroup gRPC API call.
    4. Add the load balancer to the instance group using the addTargets REST API method for the TargetGroup resource or the TargetGroupService/AddTargets gRPC API call.

  2. Check that the network load balancer named load-generator has been created and linked to the instance group:

    1. In the management console, select the folder where you created the load balancer.
    2. In the list of services, select Network Load Balancer.
    3. Click the name of the load-generator network load balancer.
    yc load-balancer network-load-balancer list

    Command result:

    +----------------------+----------------+-----------------+----------+----------------+------------------------+--------+
|          ID          |      NAME      |    REGION ID    |   TYPE   | LISTENER COUNT | ATTACHED TARGET GROUPS | STATUS |
+----------------------+----------------+-----------------+----------+----------------+------------------------+--------+
| b0ruab1ccvpd******** | load-generator | ru-central1     | EXTERNAL |              1 | b0r1tabcphde********   | ACTIVE |
+----------------------+----------------+-----------------+----------+----------------+------------------------+--------+

    Use the list REST API method for the NetworkLoadBalancer resource or the NetworkLoadBalancerService/List gRPC API call.

Create a load on an instance

  1. Get the IP address of the load balancer that you created:

    1. In the management console, select the folder containing the load balancer.
    2. In the list of services, select Network Load Balancer.
    3. Copy the IP address of the load-generator load balancer.
    yc load-balancer network-load-balancer get load-generator | grep "address"

    Command result:

      address: 84.252.133.110

    Use the get REST API method for the NetworkLoadBalancer resource or the NetworkLoadBalancerService/Get gRPC API call.

  2. Run the command to create a load:

    wrk -t20 -c20 -d20m --timeout 20s http://<load_balancer_IP_address>/sleep

    The command will launch the wrk utility that will send requests to 20 threads using 20 connections to the network load balancer for 20 minutes. The request timeout is 20 seconds. The load balancer will distribute the resulting load among the VMs from the group.

    After you launch wrk, a testing start message will appear on the screen:

    Running 20m test @ http://84.252.133.110/sleep
  20 threads and 20 connections

    Proceed to the next step without waiting for the command to complete.

Update the instance group under load

  1. In the management console, select the folder where you created the instance group.
  2. Select Compute Cloud.
  3. In the left-hand panel, select Instance groups.
  4. Click the group-for-load instance group name.
  5. Click Edit.
  6. Under Instance template, click and select Edit.
    • Under Disks, enter the new disk size of 35 GB.
    • Click Save.
  7. Click Save.
  8. Under VM states, you will see disk size changes for all VM instances in the group, step-by-step.

  1. In the specification.yaml specification, enter the new disk size of 35 GB and save the file:

    ...
size: 35G
...

  2. Update the instance group:

    yc compute instance-group update --name=group-for-load --file=specification.yaml

    Command result:

    done (9m24s)
id: cl10kktrjcn8********
folder_id: b1g7gvsi89m3********
...
   name: load-generator
service_account_id: ajehbk07uus3********
status: ACTIVE

  1. In the specification, enter the new disk size of 35 GB:

    ...
size: 35G
...

  2. To update the load-generator instance group based on the new specification, use the updateFromYaml REST API method for the InstanceGroup resource or the InstanceGroupService/UpdateFromYaml gRPC API call.

Stop the load and get the results

Stop wrk by pressing Ctrl + C.

Command result:

  Thread Stats   Avg      Stdev     Max   +/- Stdev
    Latency    10.00s     1.70ms  10.01s    68.21%
    Req/Sec     0.00      0.00     0.00    100.00%
  1148 requests in 17.47m, 211.89KB read
Requests/sec:      1.09
Transfer/sec:     206.94B

If there is no error line, e.g., Socket errors, in the result, it means all requests were processed.

How to delete the resources you created

To delete the created resources:

  1. Delete the load balancer:
    1. In the management console, select the folder you created the load balancer in.
    2. In the list of services, select Network Load Balancer.
    3. To the right of the load-generator line, click and select Delete.
    4. In the window that opens, click Delete.
  2. Delete the instance group:
    1. In the management console, select the folder where you created the instance group.
    2. In the list of services, select Compute Cloud.
    3. In the left-hand panel, select Instance groups.
    4. To the right of the load-generator line, click and select Delete.
    5. In the window that opens, click Delete.
  3. Delete the service account:
    1. In the management console, select the folder where you created the service account.
    2. In the list of services, select Identity and Access Management.
    3. In the left-hand panel, select Service accounts.
    4. In the list that opens, click and select Delete in the row with the yc-auto-sa service account.
    5. In the window that opens, click Delete.
  4. Delete the network and subnets:
    1. In the management console, select the folder where you created the network and subnets.
    2. In the list of services, select Virtual Private Cloud.
    3. Select the yc-auto-network network.
    4. Under Subnets:
      1. To the right of the yc-auto-subnet-1 line, click and select Delete.
      2. In the window that opens, click Delete.
      3. Delete the yc-auto-subnet-2 subnet as well.
    5. In the top-right corner, click Delete.

Run the following commands one by one:

yc load-balancer network-load-balancer delete load-generator
yc compute instance-group delete load-generator
yc iam service-account delete yc-auto-sa
yc vpc subnet delete yc-auto-subnet-1
yc vpc subnet delete yc-auto-subnet-2
yc vpc network delete yc-auto-network
  1. Delete the load-generator load balancer using the delete REST API method for the NetworkLoadBalancer resource or the NetworkLoadBalancerService/Delete gRPC API call.
  2. Delete the load-generator instance group using the delete REST API method for the InstanceGroup resource or the InstanceGroupService/Delete gRPC API call.
  3. Delete the yc-auto-sa service account using the delete REST API method for the ServiceAccount resource or the ServiceAccountService/Delete gRPC API call.
  4. Delete the yc-auto-subnet-1 and yc-auto-subnet-2 subnets using the delete REST API method for the Subnet resource or the SubnetService/Delete gRPC API call.
  5. Delete the yc-auto-network network using the delete REST API method for the Network resource or the NetworkService/Delete gRPC API call.
Previous
Autoscaling a group of VMs to process messages from a queue
Next
Getting started with Packer